Love Big Hero 6? The movie wouldn’t have been possible without Disney’s new Hyperion rendering tech.
If you follow our Movies & TV coverage, you already know that we love Disney’s recent Big Hero 6 — and if you’ve seen it, you probably love it, too. But the movie’s more than a great story and gorgeous animation: Big Hero 6 is a big technical breakthrough for Walt Disney Animation Studios, using a new rendering system called Hyperion that allows them to really scale up on the number of things that can go in a single shot.
And this isn’t just behind-the-scenes tech: Hyperion means that in some shots you really can see all 83,000 buildings that make up the fictional town of San Fransokyo or the millions of micro-bots moving through shots. These are scenes that wouldn’t have been possible using Frozen-era rendering technology, and Disney says that the virtual worlds of
Tangled, Frozen, Wreck-It Ralph and Bolt could fit inside the world of Big Hero 6 with room to spare. So it should be a surprise that it’s the tech behind Big Hero 6 that really gives the film a tremendous sense of scale that you don’t always find in an animated — 3D or otherwise — feature.
To find out more about the making of Big Hero 6 and the tech that made it possible, we sat down with Hank Driskell, the film’s Technical Supervisor. For those who haven’t seen the film yet, there are some minor spoilers that come up in our conversation, though nothing that’s likely to ruin the movie experience.
Q: Can you tell us a bit about what you did on Big Hero 6?
A: My job is technical supervisor. I came on board with the Visual Effects Supervisor, Kyle Odermatt, in May 2012. We were pretty much the first people to come on to the film. Kyle and I share an office, we’re kind of joined at the hip.
Kyle’s job is to work with the directors and art director to figure out their vision and then supervising all the artists on the floor who are actually making everything happen. My job as technical supervisor is to figure out all the R&D: anything we’re going to change about the way we make our movies, any new tools, techniques, or processes that the film might need as well anything the studio wants to do and use this film as an excuse for. I’m shepherding all of that.
We have a central technology group which builds tools for all the artists, so I’m making sure they’re working on things that are going to help us get the movie made, making sure they’re hearing everything they need to hear to get the tools right. As well as supervising what we call technical directors (TDs). They’re software folks that are building tools in various departments for the artists. And then as we transition from pre-production — where we’re figuring out how to make the movie — we go into “don’t touch it, you’ll break it” mode when we’re just trying to get shots out the door.
During that phase I still supervise all the TDs on the floor who are working alongside the artists to make sure everything’s running. Any time the artists have issues with data or issues with a tool breaking, we’re jumping on it to get them working again. We’re all kind of team firefighter on the show, just trying keep it all going.
Q: Can you give us a the layman’s rundown of what Hyperion does?
A: For three or four years, we’ve been amping up the complexity of our films with each successive film from Bolt to Tangled to Wreck-it Ralph to Frozen. Each film has gotten more and more complex. The tools we were using, the processes we were using, we weren’t going to be able to put as much stuff on screen as we wanted to.
And then one of our principle engineers in the technology group, Brent Burley, had a really, really clever idea for a different way of rendering. He passed the idea around to some of us and we all went “oh, wow, this is really cool.” He got a small team of people and they started working on it as a little science experiment, just to try it out and see if the idea held up.
And in the meanwhile, Kyle and I had rolled on to Big Hero 6 and were starting to ramp up pre-production. We started looking at what we all wanted with the film. It was our first collaboration with Marvel, and though it was very much a Disney movie but we wanted it to be a Marvel movie. Marvel movies have this sense of scope and scale that we really wanted to capture. Don [director Don Hall] had this vision for San Fransokyo as a mash-up of San Francisco and Tokyo. We wanted it to be a really rich urban environment — both of those cities, they’re very alive. As we started to look at that, we realized we didn’t have the tools to do it. So the science experiment came to a close in early 2013 when we flipped the switch from science experiment to “let’s make a renderer!”
We made Hyperion, this whole new way of rendering, while we were making the movie, which a lot of people thought we were nuts to even try. But we had a lot of faith in Brent and the people he gathered around him. We pulled it off somehow and the imagery really speaks for it. It’s a big step up for us in the amount of stuff we can put on the screen and the richness of the images themselves. We’re really, really proud of it.
Q: What was the biggest technical hurdle to making Big Hero 6?
A: Certainly Hyperion. It was really ambitious to craft a whole new renderer while making a movie relying on that renderer to produce imagery. There was a lot of stress and strain just trying to get that up on its feet. There were a lot of features coming online as shots were waiting for those features. It was a very tense thing.
The most stressful and exciting thing on the film was putting that together. It’s a sea change in what we’re capable of. It laid the groundwork for every film that comes after, because it makes us able to do things we just couldn’t have done before. We’re really happy that we did that, we’re really happy with what it’s giving us already for the films we’re working on after this… but while we were doing it, it was really stressful.
Q: What can you tell us about the digital world of San Fransokyo created for the film?
A: It started out like everything does here: with research. We sent people on trips to Tokyo and San Francisco to take lots and lots of photos and video of everything, from the trash in alleyways to the sky in both cities. We started to figure out just what that amalgam city looks like, that combination of both.
Then we started planning out scope and scale. Geographically we decided to layer it on top of San Francisco. We really liked the rolling hills and the feeling of the bay and all of that, so we started with that. We got ahold of city data from the City of San Francisco for the price of $5. The data tells you every lot in the city, where it is, how big it is, how big the buildings are on it, the base plan of the buildings on that lot (in terms of the shape of it), how many stories tall it is, and a whole bunch of things we didn’t need, like how many bedrooms and bathrooms there are. The stuff we did use was where it is, what shape it is, how tall it is. We used that as the foundation.
There were 83,000 buildings across 23 districts and we built them all, as well as populating the city streets with 200,000 trees and over 200,000 street lights. A lot of scenes where you’re flying over the city show all of it, with tens of thousands of buildings in one shot. It’s really exciting. Without Hyperion, we never would have been able to put all of that in a single shot.
Q: Were there any particular challenges to creating an environment on that scale?
A: From my perspective, it was gathering the right people, having all we really needed, and planning it all out. Then getting the tools to make progress, from Hyperion itself, to the city-building tools we put together, to the crowd system that we built (a whole new system called Denizen) for populating those streets with people and traffic. We started assembling all these different pieces.
A lot of it was managing the social engineering aspects. Being that ambitious provokes a lot of anxiety from a lot of people, because at the end of the day we do have a deadline. We have a release date, we do have to get the movie out. A fun part of the job is riding that line of being a little crazy and not too crazy. I think we walked that line successfully on this one and we’re really proud of what we pulled off.
Q: You talked about the crowd mechanics, but what about the micro-bots. Was that a particular technical hurdle?
A: It’s not using the same process. Denizen is for crafting our human crowds, and all of our people walking around on the street. The micro-bots were a product of our effects department. The micro-bots are really, really fun from a technical standpoint of being able to put together millions of micro-bots, sometimes in one shot.
But there’s also the emotion of them. When Hiro’s using the micro-bots, they’re very different than when Yokai is using the micro-bots. They’re really a reflection of the emotions of the person using them. The effects animators who did that work really conveyed that in a wonderful way. It’s another place where we could have cheated, but if you actually look in close the micro-bots are crawling over each other in this fun, kind of creepy look. We wanted to do something different with those.
This movie is a celebration of science, a celebration of technology and how exciting and fun this stuff could be. Very much the message we wanted to convey was that technology is only bad when bad people are driving it. We tried to keep it all grounded — we didn’t want it to seem magic-y. So as the micro-bots are traveling and building things and moving things, we tried very hard to keep trellis structures underneath them so they could support the weight. We did all kinds of of these fun things to convey these could actually be real some day and they’re not just magic. They’re not just flying things that defy physics. That was an important factor that the artists had to keep in mind while delivering those shots.
Q: How was Big Hero 6 a different challenge than previous 3D animated features? Was it just the sense of scale?
A: There are a lot of things. Certainly scope and scale was a big part of it. We wanted this richer environment, we wanted the sense of spectacle you get from a movie in the Marvel Cinematic Universe. We wanted to capture some of that because we really though there was a middle ground we could reach between a classic Disney story and one of the Marvel cinematic films. We were aiming for something that was the best of both, so that definitely had a lot of challenges.
A lot of it was self-imposed. It wasn’t like we had Marvel guys breathing down our necks. We had Joe Quesada and Jeph Loeb and Brian Michael Bendis coming in for screenings and offering ideas. They were loving everything we were doing.
We’re big Marvel fans. We’re big Disney fans. Every movie we make has that sense of pressure: the name on the door, the legacy of 80 years of filmmaking. With every movie we make here, we want to make something that’s worthy of putting that name on it. It was a combination of both of those things: we wanted to do something that really felt like it lived up to the Marvel films as well as the Disney name.
